Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Qwen3.6-Plus: Towards real world agents

qwen.ai

Post reply on HN
Print view View original

211–220 of 235 posts

Re: Qwen3.6-Plus: Towards real world agents

#211

Post by rurban »

drob518 wrote: post #208
kadoban wrote: post #193

Earlier quoted context omitted.

I'm running qwen 3.5 397b on very standard hardware. Just use the unsloth quants, they're great. I get like 20t/s or something. It's super not a publicity stunt, qwen 3.5 is the base of the best local models out there IMO.

Well, you didn’t post the specs on your rig. I think it’s probably more correct to say that you run it on very beefy but readily available hardware. My point was not that nobody could run a 300B model, but rather that a 300B model is not going to be runnable by a majority of people. Sure, anyone who wants to run that model and has the money to purchase the hardware can do it. But the hardware is going to be pricey an…

I can run a 300b model, but I don't do it. We need the H100's for training

Re: Qwen3.6-Plus: Towards real world agents

#212

Post by rurban »

throwaway911282 wrote: post #186

ignoring gpt 5.4! I feel bad for people who have not even tried it. for the same 20$ I pay to openai and anthropic, I get significantly more from openai

Same for me. But I tried the gpt business account for 35€. Did beat Claude in all instances.

Re: Qwen3.6-Plus: Towards real world agents

#213

Post by Gracana »

Aurornis wrote: post #104
Gracana wrote: post #76

Earlier quoted context omitted.

I don't think there's so much naivety. People can be aware of the the plan and still be frustrated and disappointed when it happens.

For a brief moment there were a lot of comments about how Chinese tech companies are our saviors in the age of AI because they were releasing their models. It was an edgy contrarian take that was getting a lot of traction, mostly from commenters who were unfamiliar with Alibaba and thought it was the anti-Big-tech

My explanation is simpler and does not rely on assuming that anyone is an idiot. Or an edgy contrarian.

Re: Qwen3.6-Plus: Towards real world agents

#215

Post by mtrifonov »

Most agent work focuses on task completion. Browse the web, fill out the form, and/or write the code. The harder problem is social agency, where the AI has to decide whether to participate at all. We built a cheap model gate that reads the conversational dynamics of a group chat before the expensive model runs. Wonder how Qwen3.6 performs in these nuances cases.

Re: Qwen3.6-Plus: Towards real world agents

#216

Post by kadoban »

drob518 wrote: post #208
kadoban wrote: post #193

Earlier quoted context omitted.

I'm running qwen 3.5 397b on very standard hardware. Just use the unsloth quants, they're great. I get like 20t/s or something. It's super not a publicity stunt, qwen 3.5 is the base of the best local models out there IMO.

Well, you didn’t post the specs on your rig. I think it’s probably more correct to say that you run it on very beefy but readily available hardware. My point was not that nobody could run a 300B model, but rather that a 300B model is not going to be runnable by a majority of people. Sure, anyone who wants to run that model and has the money to purchase the hardware can do it. But the hardware is going to be pricey an…

I just have a 3090 and 64gb ram. Yes this is more than most people have, but calling it a "publicity stunt" is just so uncharitably weird of a characterization.

There's smaller models all the way down too.

Like this should be _exactly_ what we want companies to release.

Re: Qwen3.6-Plus: Towards real world agents

#217

Post by scottcha »

edg5000 wrote: post #172
scottcha wrote: post #53

Earlier quoted context omitted.

I switch between Claude Code (Opus/Sonnet) and Qwen (OpenCode, OpenClaw) multiple times throughout the day and Qwen 3.5 is really nice. I do also use KimiK2.5 and GLM5 pretty often too and I'm starting to get a sense that the agent tool is becoming a little more important than the model with these level of models. As long as tool calling and prompt quality is all configured correctly by the provider.

What are the reasons for switching? Personally I got into the habit of doing a bit of a round robin with Codex/Claude (CLI) and then DeepSeek and Qwen web chat. And Claude in web chat. I like to switch just to learn the differences, otherwise I'd never know what the other models can do. But I still feel attached to Opus, but this can be fammillarity. If I only had Qwen maybe it would be effectively identical at the e…

Mine are pretty unique since we optimize the energy for and run an inference service api so forces me to dogfood alot of different options.

Re: Qwen3.6-Plus: Towards real world agents

#218

Post by drob518 »

kadoban wrote: post #216
drob518 wrote: post #208

Earlier quoted context omitted.

Well, you didn’t post the specs on your rig. I think it’s probably more correct to say that you run it on very beefy but readily available hardware. My point was not that nobody could run a 300B model, but rather that a 300B model is not going to be runnable by a majority of people. Sure, anyone who wants to run that model and has the money to purchase the hardware can do it. But the hardware is going to be pricey an…

I just have a 3090 and 64gb ram. Yes this is more than most people have, but calling it a "publicity stunt" is just so uncharitably weird of a characterization. There's smaller models all the way down too. Like this should be _exactly_ what we want companies to release.

I apologize. I didn’t mean to suggest a “publicity stunt” was a negative. Perhaps I should have said that it was a great marketing strategy. My point was, they can cite all the metrics associated with a frontier model and yet to actually get those metrics most users will have to purchase cloud-based services. That all. And sure, some people will definitely be able to run the model and benefit from it. As you say, this is what we want.

Re: Qwen3.6-Plus: Towards real world agents

#219

Post by sumedh »

throwaway911282 wrote: post #186

ignoring gpt 5.4! I feel bad for people who have not even tried it. for the same 20$ I pay to openai and anthropic, I get significantly more from openai

GPT 5, 5.1 were bad which affected its reputation and most people went with Claude because it actually worked. There is no reason to switch if you are happy with Claude.

Re: Qwen3.6-Plus: Towards real world agents

#220

Post by 8n4vidtmkvmk »

jona-f wrote: post #195
Aurornis wrote: post #11

This is their hosted-only model, not an open weight model like they’ve become known for. They got a lot of good publicity for their open weight model releases, which was the goal. The hard part is pivoting from an open weight provider to being considered as a competitor to Claude and ChatGPT. Initial reactions are mostly anger from everyone who didn’t realize that the play along was to give away the smaller models as…

> Most users of cheap API tokens are not loyal to any brand In the exploration phase, yes. But once your setup settles down you likely want to stay on the same model for stable operation.

I feel like this is true. I don't mind being a blip behind the bleeding edge if I don't have to change my tooling every month. But the second my current provider tries to screw me over, I'll still jump ship
Post reply on HN